May 7, 1992 - targeted for control with lBt-toxin-producing crops. A labora- tory strain of H. virescens developed resistance in response to selection with the Bt ...
Proc. Natl. Acad. Sci. USA Vol. 89, pp. 7986-7990, September 1992
Agricultural Sciences
Broad-spectrum resistance to Bacillus thuringiensis toxins in Heliothis virescens FRED GOULD*t, AMPARO MARTINEZ-RAMIREZt, ARNE ANDERSON*, JUAN FERREt, FRANCISCO J. SILVAt, AND WILLIAM J. MOAR§ *Department of Entomology, North Carolina State University, Raleigh, NC 27695; tDepartmento de Genetica, Universitat de Valencia, 46100-Buijassot (Valencia) Spain; and IDepartment of Entomology, Auburn University, Auburn, AL 36849
Communicated by Robert L. Metcalf, May 7, 1992
ABSTRACT Evolution of pest resistance to al proteins produced by Baciuw thw'mgwensis (Bt) would dera our ability to control agricltural pests with genetically engineered crops designe to express genes coding for these proteins. Previous genetic and biochemical analyses of insect strains with resistance to Bt toxins indicate that (a) resistance is resricted to single groups of related Bt toxins, (it) deased toxin sensvity is acted with chg in Bt-toxin binding to sites in brush-border membrane vesiles of the larval midgut, and (i) resistance is inherited as a partially or fully recessive trait. If these three characteristics were common to all resistant insects, specific crop-variety deployment strates problems associated with resscould s fgn ly dim tance in field populations of pests. We present data on Bt-oxin ristance in Heliothis vescens, a major agricultural pest targeted for control with lBt-toxin-producing crops. A laboratory strain of H. virescens developed resistance in response to selection with the Bt toxin CryIA(c). In contrast to other cases of Bt-toin rance, this H. virescens strain exhibit crossresistance to Bt toxins that differ signlficantly in strutre and activity. Furthermore, the resance in this strain is not accompanied by sigifcat changes in toxin binding, and resistance is inherited as an additive trait when larvae are treated with high doses of Cry>(c) toxin. These findings have important implications for Bt-toxin-based pest control.
Indian meal moths that were >100-fold resistant to the Bt toxin CryIA(b) were not at all resistant to CryIC (12). Diamondback moths that were >200-fold resistant to CryIA(b) were not resistant to CryIB or CryIC (13). A strain of Heliothis virescens selected with CryIA(b) and a mixture of CryIA and CryIIA proteins in an HD-1 strain of Bt was significantly resistant to only some strains of Bt (14). Restriction of resistance to a single or highly related group of toxins may be explained by studies indicating that the biochemical basis for resistance includes changes in receptors in the insect midgut (12, 13, 15). While high levels of resistance are cause for concern, the specificity of resistance has led to a belief that as insects become resistant to one Bt toxin, that toxin could be successfully replaced by a different Bt toxin (16). Additionally, genetic analyses of Bt resistance have demonstrated that the resistance is usually inherited as a mostly recessive trait (14, 15, 17-20). The rate at which such recessive traits become established in a population can be significantly decreased by the proper use of Bt-toxin-producing plants, so these genetic results have prompted the development of specific "resistance management" strategies (21, 22). We report here on Bt-toxin resistance in a strain of H. virescens, a pest of cotton, soybean, tobacco, tomato, and other agricultural crops. The resistance in our strain of H. virescens is not toxin-specific, does not appear to be related to changes in midgut receptors, and is not inherited as a recessive trait when larvae are exposed to high doses of Bt
One of the few microbes that has been used successfully in agricultural insect pest control is Bacillus thuringiensis (Bt). Liquid and powder formulations of this bacterium hold a small but growing share of the pest-control-agent market (1). Genes from Bt that code for production of a set ofinsecticidal proteins are being cloned and transferred to a number of crop plants (2-4). Expression of Bt genes in crop plants is appealing because the toxicity of the proteins coded for by these genes is restricted to specific groups of insects (5) and because plants producing these toxins are not expected to interfere with the action of natural enemies of pests (e.g., refs. 6 and 7). Many toxic proteins with varying degrees of homology in
toxin.
EXPERIMENTAL PROCEDURES Design of Selection Experiment. As part of a larger study on insect resistance to Bt toxins, a strain of H. virescens was selected for survival on an artificial diet containing CryIA(c). The selected strain and the control strain were initiated from a sample of a field population collected in July 1988. Precautions, which are described below, were taken to avoid genetic bottlenecks during the initiation of the strains. The individually laid eggs of H. virescens were collected from three tobacco fields near Carpenter, North Carolina, over a period of 4 days. A total of 700 neonates from these collection sites were reared in the laboratory on artificial diet, in individual 25-ml cups (23). Adults that emerged from this F0 generation were divided equally among 25 oviposition containers. After ca. 5 days of oviposition, 2 moths from each of the oviposition containers were checked for disease. Seventy F1 larvae from each container were reared on artificial diet (i.e., a total of 1750 larvae). The first 50 larvae to pupate from each set of 70 were placed in oviposition
amino acid sequence have been isolated from worldwide collections of Bt strains (8). At least nine distinct proteins have been characterized that are toxic to lepidopteran caterpillars (8). Toxicity of these proteins is generally related to their ability to bind to receptors in the larval midgut, and it has been demonstrated that within a single insect, there may be different receptors for different Bt toxins (9-11). Excitement over the potential of engineered plants with Bt genes has been tempered by laboratory and field work which indicates that pest insects have the capacity to adapt to Bt and its toxic proteins (34). However, studies of insect resistance to Bt toxins have found that resistance is toxin-specific.
Abbreviations: Bt, Bacillus thuringiensis; BBMV, brush-border membrane vesicle. tTo whom reprint requests should be addressed at: Department of Entomology, North Carolina State University, Box 7634, Raleigh, NC 27695-7634.
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 7986
Agricultural Sciences: Gould et al. containers. This rearing regime was continued through the F3 generation to permit the field strain to adapt to laboratory conditions (see ref. 24). The selection experiment was begun in the F4 generation, using an equal number of larvae from each of the initially isolated groups. A temporally staggered control strain was developed over time so that neonate larvae were always available for testing. The total number of larvae reared per generation in this control strain was generally between 400 and 720. Larvae from the selected strain, CP73-3, were placed on a diet containing purified toxin from Bt strain HD-73, which only produces CryIA(c). Selection was carried out in each generation unless the vigor of the colony was judged to be low or egg production was insufficient. The CryIA(c) protein concentration was adjusted to between 0.15 and 0.4 pug per ml of diet, depending on the tolerance of the colony and the quality of the lot of toxin being used. Percent survival was recorded after an average of 5.4 days, and larvae were then transferred to normal diet to complete their development. Preparation of Bt Toxins. The CryIA(c) toxin was purified from HD-73 with techniques described by Hofte et al. (25). The Bt toxins tested in the cross-resistance study were CryIA(a), CryIA(b), CryIB, CryIC, and CryIIA. The CryIA(a) and CryIA(b) toxins were cloned and produced as 133-kDa and 130-kDa proteins in a recombinant Escherichia coli system described by Hofte et al. (25, 26). CryIB was cloned from Bt var. entomocidus HD-110. Its sequence is identical to that published by Brizzard and Whiteley (27). CryIC was also cloned from HD-110. It differed slightly from the CryIC gene isolated by Honee et al. (28) as described by Ferre et al. (13). These proteins were truncated to their toxic form by trypsin digestion (26). M. Peferoen and B. Lambert (Plant Genetic Systems, Ghent, Belgium) provided these proteins as a gift. The 65-kDa CryIIA protein was cloned and produced in E. coli by the method of Moar et al. (29). Its nucleotide sequence is identical to that of other cloned CryIIA genes (29). Purity of CryI and CryII proteins was judged to be >80%, based on SDS/PAGE (see ref. 29). Testing Resistance and Cross-Resistance. Degree of resistance to CryIA(c) was first tested after four generations of selection (30-40 days per generation). Tests for resistance to other toxins were conducted between generations 13 and 15 of selection. Probit analysis (30) of larval survival over a range of doses (three to six) was used for assessing degree of resistance, except in the first tests with CryIA(c), where a single dose was used, and in tests with toxins that caused 0
2-
C
1-
D
0-
20
40
60
0
CryIB, ,ug/ml 4
.0)
3-
3
2
co
co
E.
(b.C
Table 2. Concentration of binding sites (Ro) and apparent equilibrium dissociation constant (Kd) of Bt crystal proteins incubated with BBMVs of H. virescens Rt, pmol/mg of Toxin Strain vesicle protein Rt/Kd Kd, nM CryIA(c) Control 0.95 ± 0.15 3.40 ± 0.24 3.57 Selected 0.76 ± 0.17 2.90 ± 0.40 3.80 2.60 ± 1.20 CryIA(b) Control 3.30 ± 1.80 1.27 Selected 1.10 ± 0.40 1.00 ± 0.20 0.91 Kd and Rt values (mean ± SE) were calculated from homologous
2-
I
1
-
Results of this study indicate that selection of H. virescens with a single type of Bt toxin can lead to broad-spectrum Bt resistance. The finding of cross-resistance between CryIA(c) and CryIA(b) is not surprising, because these two toxins are similar in structure and toxicity. Of all the evidence regarding cross-resistance, that between CryIA(c) and CryIIA was the least expected. The amino acid sequence of CryIIA is very different from that of CryIA(c) (8, 29). Furthermore, recent studies on the mode of action of CryIIA in Helicoverpa zeal indicate that in contrast to previously studied Bt toxins, CryIIA does not show saturable binding to BBMVs. When CryIIA was incubated with BBMVs, it did not inhibit subsequent binding of CryIA(c), but when CryIA(c) was incubated with BBMVs, it inhibited the nonsaturable binding of CryII(a). Additionally, CryIIA did not lead to the voltageindependent cation-selective channels in planar lipid bilayers, characteristic of CryIA(c).f Binding studies with CryIA(c) and CryIA(b) indicated that the resistance was not due to changes in BBMV-saturable binding characteristics. The existence of strong crossresistance to CryIIA, which does not show saturable binding, would not have been expected ifthe resistance were only due to saturable binding changes. Our results contrast with those of MacIntosh et al. (15), who found that an H. virescens strain selected with CryIA(b), and with an HD-1 strain that produces multiple toxins (14, 15), had a lower affinity for CryIA(b) and more receptors per mg of protein than a susceptible strain. We found no significant differences in CryIA(b) binding characteristics of our strains, but the trend in our data was for the resistant strain to have higher affinity and fewer receptors than the susceptible strain. This information, as well as the cross-resistance data, suggest that the mechanism of resistance in our strain is different from that of the strain tested by MacIntosh et al. (15). The resistant strain tested by MacIntosh et al. (15) had 71-fold resistance to CryIA(b) and only 16-fold resistance to CryLA(c), whereas our strain was more resistant to CryIA(c) than to CryIA(b).
CrylC, jg/ml
FIG. 2. Growth of larvae of the control (-) and selected (A) H. virescens strains exposed to various concentrations of CryIA(a) for 8 days (A), CryIB for 8 days (B), or CryIC for 6 days (C).
lEnglish, L., Robbins, H. L. & Slatin, S., First International Con-
ference on Molecular Biology of Bacillus thuringiensis, eds. Aronson, A., Bulla, L., Carlton, B. & Rapaport, G., July 26-29, 1991, San Francisco, p. 19 (abstr.).
Proc. Natl. Acad. Sci. USA 89 (1992)
Agricultural Sciences: Gould et al.
A
-
80-
.?
4. II * 4
o 60-
E
I
4
C
20 a) a.
I*
......s* Control strain
40. 4.
Control 9 x selected d -@-@-@- Selected 9 x control d - - - - Selected strain
4
40.
20 -
40. 0.01
4.P
100
10
1
0.1
100
B
.:
4.
E
/
40-
I
4. 4.
/
/
0 60C
4.
-/
80
Control strain F1 crosses Selected strain
4.
a.
4.
4. 4. 20
7989
-
.4. . . .. .
I.
I
^ .1 0.1 1
.1 0.01 -1
I
-X, 10 100
CrylA(c), /ug/ml FIG. 3. Percent mortality of the control and selected strains as well as progeny from crosses as a function of the CryIA(c) concentration (note loglo scale) in their diet. (A) First cross with reciprocal crosses tested separately. Mortality lines for the progeny of the reciprocal crosses are nearly identical. (B) Second cross with reciprocals tested together.
Results of the genetic crosses indicate that at toxin concentrations that kill only 10-20% of the susceptible line, resistance is at least partially recessive. However, at higher concentrations that kill 80-90% of the susceptible strain, 100
A I
0*
8060-
E
40-
E 20I E
.
0.1
.
1
.
10
.
.
100
1000
ONO B
1004 m
60 -
resistance is inherited as an additive trait. If plants are engineered to produce enough toxin to kill most of the susceptible larvae and are planted over large areas, response to selection would be rapid if larvae with such intermediate Bt-toxin tolerance were present (21). We have not yet determined the number of loci involved in resistance. Previous studies of Bt-toxin resistance found that resistance was restricted to a single type of toxin. This led to a hypothesis that it would be possible to replace one Bt toxin with another as insects adapted to specific toxins (12, 16). Our findings clearly indicate that selection with a single Bt toxin could lead to broadly based resistance that would preclude control of an insect population with any Bt product. There are now two cases in which resistance to Bt has been clearly shown to be restricted to a single class of Bt toxins (12, 13) and one case in which resistance is not specific (this study). More studies of cross-resistance patterns and of the mechanisms of resistance will be needed before any generalizations can be made. Until then, we cannot assume that discoveries of new Bt toxins with unique biochemical properties will decrease the consequences of misusing Bt.
80
6040 ^ ..--
......I
0.1
1
......
10 Competitor, nM
... .0..
100
We are grateful to M. Peferoen and B. Lambert of Plant Genetic Systems for freely supplying us with purified toxins. R. Hampton, D. Ashby, D. Bumgarner, L. Bumgarner, S. Goodwin, and S. Braxton provided excellent technical help. Financial support was provided by grants from the United States Department of Agriculture, CIBAGeigy, the North Atlantic Treaty Organization, the European Community (ECLAIR), and Consellaria Valenciana de Educacio i Ciencia.
1000
FIG. 4. (A) Binding of 1251-labeled CryIA(c) to BBMVs at various concentrations of nonlabeled CryIA(c) competitor. (B) Binding of 125I-labeled CryIA(b) at various concentrations of nonlabeled CryIA(b). *, Susceptible strain; o, resistant strain.
1. Gelernter, W. D. (1990) in Managing Resistance to Agrochemicals, ACS Symposium Series 421 (Am. Chem. Soc., Wash-
ington). 2. Gasser, C. S. & Fraley, R. T. (1989) Science 244, 1293-1299. 3. Meeusen, R. L. & Warren, G. (1989) Annu. Rev. Entomol. 34, 373-381.
7990
Agricultural Sciences: Gould et al.
4. Perlak, F. J., Fuchs, R. L., Dean, D. A., McPherson, S. L. & Fischhoff, D. A. (1991) Proc. Natl. Acad. Sci. USA 88, 33243328. 5. Van Rie, J. (1991) Trends Biotechnol. 9, 177-179. 6. Johnson, M. T. & Gould, F. (1992) Environ. Entomol. 21, 586-597. 7. Thoms, E. M. & Watson, T. F. (1986) J. Invert. Pathol. 47, 178-183. 8. Hofte, H. & Whiteley, H. R. (1989) Microbiol. Rev. 53, 242255. 9. Hoffman, C., Luthy, P., Hutter, P. & Pliska, V. (1988) J. Biochem. (Tokyo) 173, 85-91. 10. Van Rie, J., Jansens, S., Hofte, H., Degheele, D. & Van Mellaert, H. (1989) Eur. J. Biochem. 186, 239-247. 11. Wolfersberger, M. G. (1990) Experientia 46, 475-477. 12. Van Rie, J., McGaughey, W. H., Johnson, D. E., Barnett, B. D. & Van Mellaert, H. (1990) Science 247, 72-74. 13. Ferre, J., Real, M. D., Van Rie, J., Jansens, S. & Peferoen, M. (1991) Proc. Natl. Acad. Sci. USA 88, 5119-5123. 14. Marrone, P. G. & MacIntosh, S. C. inBacillus thuringiensis:An Environmental Biopesticide, eds. Entwistle, P. F., Cory, J. S., Bailey, M. J. & Higgs, S. R. (Wiley, New York), in press. 15. MacIntosh, S. C., Stone, T. B., Jokerst, R. S. & Fuchs, R. L. (1991) Proc. Natl. Acad. Sci. USA 88, 8930-8933. 16. Cutler, K. (1991) Ag. Biotech. News Jan./Feb., p. 7. 17. McGaughey, W. H. & Beeman, R. W. (1988) J. Econ. Entomol. 81, 28-33. 18. Sims, S. R. & Stone, T. B. (1991) J. Invert. Pathol. 57, 206210.
Proc. Nad. Acad. Sci. USA 89 (1992) 19. Martinez-Ramirez, A. C., Escriche, B., Real, M. D., Silva, F. J. & Ferre, J., Pestic. Sci., in press. 20. Tabashnik, B. E., Schwartz, J. M., Finson, N. & Johnson, M. W. (1992) J. Econ. Entomol. 85, 1046-1055. 21. Gould, F. (1986) Environ. Entomol. 15, 1-10. 22. Gould, F. (1988) BioScience 38, 26-33. 23. Burton, R. L. (1970) J. Econ. Entomol. 63, 1969-1970. 24. Service, P. M. & Rose, H. A. (1985) Evolution 39, 943-945. 25. Hofte, H., de Greve, H., Seurinck, J., Jansens, S., Mahillon, J., Ampe, C., Vandekerckhove, J., Vanderbruggen, H., van Montagu, M., Zabeau, M. & Vaeck, M. (1986) Eur. J. Biochem. 161, 273-280. 26. Hofte, H., Van Rie, J., Jansens, S., Van Houtven, A., Vanderbruggen, H. & Vaeck, M. (1988) Appl. Environ. Microbiol. 54, 2010-2017. 27. Brizzard, B. L. & Whiteley, H. R. (1988) Nucleic Acids Res. 16, 4168-4169. 28. Honee, G., van der Salm, T. & Visser, B. (1988) Nucleic Acids Res. 16, 6240. 29. Moar, W. J. (1990) Ph.D. dissertation (Univ. of California, Riverside). 30. SAS Institute (1985) SAS Users Guide: Statistics (SAS Inst., Cary, NC). 31. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254. 32. Munson, P. J. & Rodbard, D. (1980) Anal. Biochem. 107, 220-239. 33. Tabashnik, B. E., Cushing, N. L., Finson, N. & Johnson, M. W. (1990) J. Econ. Entomol. 83, 1671-1676.
